Newsgroups: comp.robotics
From: pcats@cryton.demon.co.uk (Patrick Arnold)
Path: brunix!sgiblab!darwin.sura.net!howland.reston.ans.net!pipex!demon!cryton.demon.co.uk!pcats
Subject: Re: Help needed on Sonar Ranging Modules 
Distribution: world
References: <2hkqgi$4dc@netnews.upenn.edu>
Organization: The Serial Port
Reply-To: pcats@cryton.demon.co.uk
X-Newsreader: Simple NEWS 1.90 (ka9q DIS 1.21)
Lines: 70
Date: Thu, 20 Jan 1994 19:10:20 +0000
Message-ID: <759093020snz@cryton.demon.co.uk>
Sender: usenet@demon.co.uk

In article <2hkqgi$4dc@netnews.upenn.edu> saha@seas.upenn.edu writes:

>  Three of us are trying to build a robot that will be able to detect if
>something is in its path.  We are thinking of using two sonar ranging 
>modules to sweep the path ahead of the robot.  We are planning on using 
>TI's SN 28827.  
>  The problem is, we don't have too much idea about the transducer 
>that the module uses (i.e. frequency, biasing needs,..).  If there is anyone 
>with the knowledge of this module or addresses of suppliers please help us 
>out.
>  I have already called Polaroid and they only supply in bulk quantities.

The polaroid ultrasonic transducer is a capacitive high voltage device. As
memory serves, the thing uses pulses of about 300v or so, and is usually used
in applications that output a chirp, rather that a single frequency pulse. It
is very simple in construction, being basically an aluminum plate with 
concentric, very fine grooves etched into it, which is covered with a very
thin gold foil, insulated from the plate. The plate and the foil act as the
two electrodes of a capacitor.

The sensors are used in polaroid autofocus cameras, and also in a number of
ultrasonic rangerfinders, such as electronic tape measures, etc.

I have played with such a ranging system, out of a fairly old camera 
(late 70's, early 80's), and it was remarkably accurate. The advantage of a 
chirp rather than a pulse is that one or more of the emitted frequencies is
almost bound to reflect off any type of surface with sufficient amplitude to
be detected. Unfortunately I managed to blow up the electronics of the sensor
system one day, and I have never been able to find another camera cheap
enough to rip to pieces. I have also been unable to ever lay hands on the 
TI sonar module you mention, although I have tried for years.

As a slight aside, there is an alternative to sonar for object detection.
I came up with an optical system that does this quite well some years ago.
The system may well have been thought of by lots of other people, but I have
never seen it used elsewhere. It works like this:

Mount a sensitive photosensor at one end of a short opaque tube, to restrict
its field of view a lot. Lenses are optional. This tube is mounted pointing
straight ahead, at one side of the robot. At the other side, an assembly
containg a laser diode, motor, and shaft encoder is mounted, at the same 
height as the sensor. The motor can traverse the laser output horizontally in
an arc, and the shaft encoder allows the exact angular position of the beam
to be measured.

The distance from the sensor to the laser is known, and the angle of the sensor
to the body of the robot is a right angle. If the laser beam is traversed 
until the sensor detects the maximum intensity, the output spot is now 
directly in front of the sensor. This forms a triangle between the sensor, 
laser, and beam spot. The angle of the laser beam can be measured accurately
by means of the shaft encoder, and you now have all the data necessary to 
calculate the length of the other two sides of the triangle, one of which is
the distance to the target.

The entire sensor array can be mounted on a rotatable and/or tiltable 
platform, which allows the sensing point to be scanned in any direction. 

There are a number of ways to scan the laser beam, and measure its angle. I
have built systems that use rotating mirrors, and measure the angle by timing
from a fixed reference sensor, etc. Modulating the laser beam can also be 
useful, as spurious reflections, lights, etc, can then be masked out.

Hope this is useful.

Patrick.

-- 
The shortest distance between two points has not yet been built.

pcats@cryton.demon.co.uk vox+44(749)670058 fax+44(749)670809 dat+44(749)670030
